The Development of Simultaneous Congkak Game with Intelligent Agents
Nik Abdul Muhaimin Nik Ahmad Fazlan, Chye Cheah Tan
Abstract
Nik Abdul Muhaimin Nik Ahmad Fazlan, Chye Cheah Tan
Abstract
Congkak is a two-player game of perfect information with two phases, a simultaneous phase and a sequential phase. In this work, several intelligent agents for each phase were implemented to determine that minimax is the best algorithm for the sequential phase of the game. From the agents implemented, the payoff matrix of the game can be determined. It is shown that there is no pure strategy Nash Equilibrium but there exists a mixed strategy Nash Equilibrium of (0, 0, 0.111, 0.111, 0.333, 0.111, 0.333). An agent using this mixed strategy was proven to be more effective than the simultaneous agents trained.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Congkak is a two-player game of perfect information with two phases, a simultaneous phase and a sequential phase. In this work, several intelligent agents for each phase were implemented to determine that minimax is the best algorithm for the sequential phase of the game. From the agents implemented, the payoff matrix of the game can be determined. It is shown that there is no pure strategy Nash Equilibrium but there exists a mixed strategy Nash Equilibrium of (0, 0, 0.111, 0.111, 0.333, 0.111, 0.333). An agent using this mixed strategy was proven to be more effective than the simultaneous agents trained.
Key concepts: Nash equilibrium, Normal-form game, Minimax, Strategy, Best response, Stochastic game, Computer science, Repeated game